278 APPENDIX. 



Asia ; yet, in consequence of their size and the abundance of individuals, perhaps more 

 conspicuous and equally prevalent in tropical America. There are about 100 genera 

 and 900 species ; and of these fifteen and 115 respectively are known to inhabit 

 Mexico and Central America. But these numbers are probably considerably below those 

 actually existing ; that for species especially, as we record only ten from Nicaragua and 

 fourteen from Panama. Porphyrospatha, a genus of two described species, is restricted 

 to Costa Eica, and the only one endemic within our limits. On the other hand, ninety- 

 nine of the species have not been recorded from elsewhere. Philodendron and Anthu- 

 rium, both exclusively American, the former numbering about 120 species, the latter 

 160, are the most conspicuous and pervading genera throughout tropical America. 

 Many of the species are lofty climbers, reaching the tops of the tallest trees. Other 

 large forms are the species of Xanthosoma and Monstera ; but these are all surpassed in 

 size of leaf and inflorescence by the Nicaraguan Dracontium gigas (syn. Godwinia gigas), 

 which has a spathe two feet long, though this, again, is small in comparison with the 

 Sumatran Amorphophallus titanum, which has leaves twenty-five feet, and spadices six 

 feet long. 



Graminece. 



Reference has already been made (vol. iii. p. 475) to Fournier's < Monograph of the 

 Mexican Grasses,' in which a much narrower view of the limits of genera and species 

 is taken than in this work. So great indeed is the difference, that his totals of genera 

 and 'species for Mexico alone considerably exceed ours for Mexico and Central America 

 together. He recognizes 123 genera, sixteen of them being endemic, and 643 species, 

 371 of which he estimated to be endemic. It is unnecessary to repeat our reasons for 

 not accepting the greater part of his proposed new species. Had we included them in 

 our calculations they would, from our standpoint, have greatly exaggerated the endemic 

 element, as will be perceived from the following numbers : — genera ninety-nine, endemic 

 six ; species 520, endemic 265. There is no doubt, however, that these numbers will 

 be very considerably augmented by future investigations in the southern part of our area, 

 as, for instance, in Costa Eica, whence we have only nine species. 



The grasses of (Ersted's Costa Eica collection have not been elaborated ; and it appears 

 to have been the practice to distribute the duplicates as the plants were published. 

 From the Table, p. 199, it will be seen that we have worked out the distribution of the 

 two great "series" of grasses — the Panicacese, distinguished by the spikelet being 

 articulated with the pedicel below the outer glumes, and the Poacese, in which the 

 articulation is above the outer glumes ; but the results do not demand any special 

 remarks in this place. In addition to the six endemic genera, twenty-six others are 

 restricted to America, and no fewer than fifty-nine are of wide distribution. Of the non- 

 endemic species 214 are restricted to America, which, added to the endemic, leaves a 

 residue of forty-one that extend beyond America. Adding the Mexican genera and 

 species which are not localized to the South Mexican, we have a total of ninety-one 



